Automatic Quantum Computer Programming: A Genetic Programming Approach

158 Pages·2007·2.297 MB·English

by  Lee Spector (eds.)

AUTOMATIC QUANTUM
COMPUTER PROGRAMMING
A Genetic Programming Approach
GENETIC PROGRAMMING SERIES
Series Editor
John Koza
Stanford University
Also in the series:
GENETIC PROGRAMMING AND DATA STRUCTURES: Genetic
Programming + Data Structures = Automatic Programming! William B.
Langdon; ISBN: 0-7923-8135-1
AUTOMATIC RE-ENGINEERING OF SOFTWARE USING
GENETIC PROGRAMMING, Conor Ryan; ISBN: 0-7923-8653-1
DATA MINING USING GRAMMAR BASED GENETIC
PROGRAMMING AND APPLICATIONS, Man Leung Wong andKwong
Sak Leung; ISBN: 0-7923-7746-X
GRAMMATICAL EVOLUTION: Evolutionary Automatic Programming
in an Arbitrary Language, Michael O 'Neill and Conor Ryan; ISBN: 1-4020-
7444-1
GENETIC PROGRAMMING IV: Routine Human-Computer Machine
Intelligence, John R. Koza, Martin A. Keane, Matthew J. Streeter, William
Mydlowec, Jessen Yu, Guido Lanza; ISBN: 1-4020-7446-8
GENETIC PROGRAMMING THEORY AND PRACTICE, edited by
RickRiolo, Bill Worzel; ISBN: 1-4020-7581-2
AUTOMATIC QUANTUM
COMPUTER PROGRAMMING
A Genetic Programming Approach
Lee Spector
Hampshire College
^ Springer
Lee Spector
Hampshire College
Library of Congress Control Number: 2006931640
ISBN-10: 0-387-36496-X e-ISBN-10: 0-387-36791-8
ISBN-13: 978-0387-36496-4 e-ISBN-13: 978-0387-36791-0
© 2004 by Springer Science-nBusiness Media, LLC
Paperback Edition 2007
All rights reserved. This work may not be translated or copied in whole or in part
without the written permission of the publisher (Springer Science + Business
Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief
excerpts in connection with reviews or scholarly analysis. Use in connection with
any form of information storage and retrieval, electronic adaptation, computer
software, or by similar or dissimilar methodology now known or hereafter
developed is forbidden.
The use in this pubhcation of trade names, trademarks, service marks and similar
terms, even if they are not identified as such, is not to be taken as an expression
of opinion as to whether or not they are subject to proprietary rights.
987654321
springer.com
Contents
Preface vii
1. THE POWER OF QUANTUM COMPUTING 1
1 What is Quantum Computing? 1
2 Possibilities Count 4
3 The Role of Automatic Programming 10
2. QUANTUM COMPUTER SIMULATION 13
1 Bits, Qubits, and Gates 13
2 Gate-Level Simulation 18
3. QUANTUM COMPUTER PROGRAMMING 23
1 QGAME: Quantum Gate and Measurement Emulator 23
2 Visualization 28
3 Example: Grover's Database Search Algorithm 31
4. GENETIC AND EVOLUTIONARY COMPUTATION 37
1 What is Genetic and Evolutionary Computation? 37
2 Genetic Algorithms 39
3 Scalability via Parallelism 41
4 Applicability of Genetic and Evolutionary Computation 41
5. GENETIC PROGRAMMING 43
1 Programming by Genetic Algorithm 43
2 Traditional Program Representations 45
3 Traditional Genetic Operators 48
4 Example: Symbolic Regression 49
5 Obtaining Genetic Programming Results 52
vi AUTOMATIC QUANTUM COMPUTER PROGRAMMING
6. EVOLUTION OF COMPLEX PROGRAMS 55
1 Types, Modules, and Development 56
2 The Push Programming Language 59
3 Push Examples 63
4 PushGP: Genetic Programming with Push 68
5 Autoconstructive Evolution 71
7. EVOLUTION OF QUANTUM PROGRAMS 75
1 Program Representations 75
2 Fitness 80
3 Operators and Refinements 82
8. EVOLVED QUANTUM PROGRAMS 87
1 The 1-bit Deutsch-Jozsa (XOR) Problem 88
2 Grover's Database Search Problem 95
3 Scaling Majority-ON 104
4 The OR and AND/OR Problems 108
5 Gate Communication Problems 114
6 Significance of These Results 118
9. CONCLUSIONS AND PROSPECTS 123
Appendix A
QGAME source code 127
References 141
About the Author 149
Index 151
Preface
This is a book about the frontiers of computer science that have re
cently been opened by work in quantum mechanics, but it is also a book
about the use of recently developed automatic programming technolo
gies to explore those frontiers. The automatic programming technologies
themselves issue from another interdisciplinary frontier of computer sci
ence — one born of the intersection of computer science with evolution
ary biology. So this is a book about two frontiers of computer science,
one being used primarily for the sake of exploring the other.
The selection of topics in this book was made with the intention of
showing how genetic programming can be usefully applied to certain
problems in quantum computing. To this end, it provides a basic intro
duction to quantum computing for non-physicists and it also provides a
basic introduction to genetic programming for non-computer-scientists.
These treatments should be comprehensible to scientifically literate read
ers who have, at minimum, a passing familiarity with undergraduate-
level computer science (e.g. programming concepts) and mathematics
(e.g. simple linear algebra). No background in physics is assumed.
Neither the introduction to quantum computing nor the introduction
to genetic programming is intended to be comprehensive or even "bal
anced." Coverage of each field is limited to relatively narrow slices that
support the demonstrations found later in the book — those demon
strations show how certain genetic programming techniques can be ap
plied to certain problems in quantum computing. Citations are provided
where appropriate to sources that provide more comprehensive and de
tailed coverage.
The first chapter contains an introduction to quantum computing for
non-physicists. The intention is to provide readers with a sense of how
quantum computers could possibly deliver the surprising benefits that
many researchers envision.
viii AUTOMATIC QUANTUM COMPUTER PROGRAMMING
The second chapter details a mathematical (matrix-based) model of
quantum computation and describes how this model can be used to sim
ulate quantum computations on classical computers. Such simulation
is necessarily inefficient — if we could simulate quantum computers ef
ficiently on classical computers then there'd be little reason to study
quantum computing in the first place! But for small computations sim
ulation is indeed possible; this model allows us to use simulation in the
"fitness assessment" step of a genetic programming algorithm, described
later in the book.
The third chapter describes one particular quantum computer simula
tion system, the author's QGAME ("Quantum Gate and Measurement
Emulator") system, and presents a few of the ways in which quantum
programs and quantum computer states can be visually displayed. It
concludes with a detailed example of the simulation of a quantum pro
gram for Grover's database search problem.
The fourth chapter introduces genetic and evolutionary computation,
with a focus on the traditional genetic algorithm. It also discusses, in
general terms, the use of parallelism to scale genetic and evolutionary
computation technologies up for complex applications, and the applica
bility of these technologies for various types of problems including those
related to quantum computing.
The fifth chapter specializes the treatment of genetic algorithms to ge
netic programming, which is the use of genetic algorithms for automatic
programming. It includes a detailed example and a discussion of the
steps one must generally take to obtain and understand useful results
from a genetic programming system.
The sixth chapter moves beyond traditional genetic programming,
and describes the ways in which one can evolve programs that include,
for example, multiple data types, modules, and developmental compo
nents. Some of these capabilities are particularly useful for the evolu
tion of quantum programs. Emphasis is placed on the author's Push
programming language for genetic and evolutionary computation, which
provides some of the desired advanced capabilities in unusually simple
ways. This chapter concludes with a description of the PushGP genetic
programming system, which evolves Push programs, and a brief descrip
tion of some more radically self-adaptive "autoconstructive evolution"
techniques that are enabled by Push.
The seventh and eighth chapters bring the materials from all of the
preceding chapters together, first with a discussion of specific strate-
PREFACE ix
gies for quantum program evolution/ and then with concrete examples
in which interesting quantum programs were evolved using QGAME,
PushGP and related technologies. These examples document a few spe
cific ways in which genetic programming has already helped to explore
the power of quantum computing.
The ninth chapter provides a brief summary of the main points of the
book and discusses prospects for new discoveries made with the aid of
automatic quantum computer programming technologies.
Source code, in Common Lisp, for a minimal version of QGAME is
included in the Appendix. Additional related source code is available
online from addresses that are cited within the text. Most of these files
are also hnked to the author's pubhc "code" page.^
This book would not have been possible without the close working re
lationships enjoyed by the author with colleagues and students at Hamp
shire College in Amherst, Massachusetts. Several of the results that are
used as examples in the book emerged from joint work of the author
with Herbert J. Bernstein, Howard Barnum, and Nikhil Swamy. Al
though specific joint results are acknowledged where they occur in the
text, these citations do not by themselves fully convey the extent of the
influence of these colleagues. Similarly, the novel technologies that are
described in the text owe much to the contributions of Chris Perry, Jon
Klein, Mark Feinstein, Raymond Coppinger, Alan Robinson, Raphael
Crawford-Marks, and Manuel Nickschas. Many of these colleagues also
commented on the manuscript of this book, leading to substantial im
provements. Additional substantial comments were provided by John
Koza, Sameer H. Al-Sakran, and Rennie Nelson. Rebecca S. Neimark
provided essential assistance in many phases of the project, including
the creation of several of the figures and the design of the cover, which
uses an image created by Chris Perry. James Hendler provided critical
encouragement and advice, and Leni Bowen and Paula Harmon provided
invaluable administrative support.
Some of the materials used in this book derive from those prepared
by the author for a series of tutorials on quantum computing presented
over several years at the Genetic and Evolutionary Computation Confer-
•'Note that the term "evolution" is used here and throughout this book in a sense derived
from its biological usage: it refers to a process in which a population undergoes variation and
natural selection. Some physicists use "evolution" in a more general sense, to describe any
change in a system over time. The phrase "quantum program evolution" in this book refers
to the generation of quantum programs using techniques derived from biological evolutionary
processes.
^http://hampshire.edu/lspector/code.html